Five patent families, three technical layers

A metabolically orthogonal chassis, an optogenetic septal pore valve, and a scaffold geometry built for hyphal growth. Covered by US provisional application 64/087,071, filed June 10, 2026.

Layer 1 — Chassis

Metabolically Orthogonal N. crassa

The iJDZ836 genome-scale metabolic model (836 genes, ~1,200 reactions) describes a metabolically capable organism. Flux-balance analysis confirms xylose as a preferred carbon source and the valid chassis feed for the chow reservoir.

The carbon source screen found 24 of 78 tested sources support growth. Gene essentiality analysis flagged ~130 of 836 genes (~15.5%) as essential — the hard boundary for knockout targeting.

We screened 10,440 double-knockout pairs from a 145-gene pool of transporter and central carbon genes, ranking each by Composite Orthogonality Score. Sixteen strains clear the chip-integration threshold (RBS ≥ 0.868) — the recommended chassis candidates.

Orthogonality is strong but partial, and we state it that way deliberately.

The leading chassis blocks roughly 87% of reagent uptake while sustaining xylose growth at approximately 18–19% of wild-type rate. The tradeoff between blocking and fitness is a design constraint, not a defect: the chip-integration threshold is set where enough residual growth remains for robust colonization.

Phenotype Tiers

Tier Ranks Characteristics
1 — Best 1–16 Best sugar blocking, ~18–19% WT xylose growth
2 — Balanced 17–32 More aggressive blocking, marginal xylose
3 — Fitness 33–50 Best xylose fitness, weaker sugar blocking

Model

iJDZ836

KO Pairs Screened

10,440

Chip-Eligible Strains

16

Metabolic Orthogonality

Xylose ✓ Metabolized Xyl pathway Growth ~18–19% WT Glucose uptake substantially reduced Reagent metabolites uptake substantially reduced Xylose pathway active ~87% of reagent uptake blocked

IP Coverage

HIGH Composition of matter — specific genotype
HIGH Six-gene KO screening method
HIGH Δcre-1 + Δwc-1 six-gene chassis
How the valve works

The pore array already exists in the organism. The actuator adds an anchor at the rim and a light-gated plug — nothing has to be fabricated into the channel.

Three-stage mechanism of the optogenetic septal pore valve Stage one: a hypha is partitioned by septa, each with a central pore about 150 nanometres across. Stage two: in the dark, the SPA1-ePDZ anchor sits constitutively at the pore rim while the LOV2-Jα-HEX1 plug protein floats free in the cytoplasm, so the pore stays open. Stage three: blue light at 450 nanometres undocks the Jα helix, exposing a binding motif that is captured by the anchor, recruiting HEX1 to oligomerize into a plug that occludes the pore. 1 · THE ARRAY ALREADY EXISTS Hyphae are partitioned by septa. Each carries a central pore ~150 nm across, natively gated by Woronin bodies built from HEX1. A dense valve array — just not addressable. 2 · DARK — ANCHOR SET, PORE OPEN CompB — SPA1-ePDZ, constitutive at the rim CompA_v3 — LOV2-Jα-HEX1, free, Jα docked With Jα docked against LOV2, the binding motif is hidden. The plug protein stays cytoplasmic and cytoplasm moves freely between compartments. No light means no plug — fail-safe open. 3 · 450 nm — RECRUITED, OCCLUDED 450 nm Light undocks Jα and exposes the motif. The anchor captures it, and HEX1 oligomerizes at the anchored site into a plug that occludes the pore. Thermal reversion reopens it with no input.

This diagram shows the intended mechanism, not a measured result. Stage 1 is established N. crassa biology. Stages 2 and 3 are the behavior our constructs are designed to produce; whether they do is precisely what Gates 1 and 2 determine.

Layer 2 — Valve

Optogenetic Septal Pore Valve

Hyphae of N. crassa are partitioned by septa, each perforated by a central pore roughly 150 nm across, natively gated by Woronin bodies built from the self-assembling protein HEX1. The organism has already fabricated a dense valve array — it is simply not addressable.

Our two-component actuator makes it addressable. CompA_v3 fuses the Avena sativa LOV2 photosensor, through its Jα helix, to HEX1. CompB anchors an ePDZ domain at the pore via SPA1. Blue light at 450 nm undocks Jα, exposes the binding motif, and drives recruitment and oligomerization at the anchored site. Removing light reverses it — fail-safe open logic, since no light means no plug.

Component A — LOV2-HEX1 Plug

LOV2 + Jα — blue-light photosensor from Avena sativa phototropin 1. Uses ubiquitous FMN as its chromophore — no foreign cofactor required. N-terminal placement keeps the HEX1 self-assembly contacts free. pLDDT = 91 (ColabFold).

HEX1 core (150 aa) — N. crassa Woronin-body protein, naturally evolved to plug septal pores under stress. Fused C-terminal via a (GGGGS)×3 linker; freely cytoplasmic until light-tethered. pLDDT = 88.

Component B — SPA1-ePDZ Anchor

SPA1 (aa 1–146) — N. crassa septal-pore-rim localizing domain. Constitutively anchored at the pore rim; no light dependence in localization.

ePDZ (engineered PDZ, iLID system) — light-gated affinity switch: ~130 nM in the lit state (Jα exposed) vs. >10 µM dark (Jα docked, epitope hidden). pLDDT = 88.

Thermal reversion: LOV2 dark recovery (seconds–minutes) automatically reopens the valve without any input signal. True fail-safe open behavior.

What is established and what is not

Both constructs are designed, codon-optimized, and modeled. Per-domain structure confidence is high (pLDDT 91 for LOV2, 88 for HEX1).

Interface confidence is not. Our construct presents the binding motif through an internal linker rather than at a free C-terminus, departing from canonical PDZ geometry, and multimer prediction of the lit-state complex returned an inconclusive ipTM of 0.18. Two measurements resolve this — SEC-MALS and BLI — and they are the next work we do.

Valve Mechanism

DARK — Flow OPEN

≥10×

Dark/lit contrast — Gate 2 criterion

450 nm

Activation wavelength

IP Coverage

HIGH LOV2-Jα-HEX1 + SPA1-ePDZ constructs
HIGH Method of optogenetic flow control

Layer 3 — Scaffold

H-Tree Scaffold & Integration

The H-tree topology is the critical design choice. Every one of the 16 terminal reagent ports is exactly the same path length from the central chow reservoir — ensuring simultaneous, pressure-balanced reagent delivery.

Channel Sizing Rationale

Level Width Hyphae Function
Root–L1 20 µm 4–8 Main trunk — flow redundancy
L2–L3 13–17 µm 2–3 Valve control zone
Depth (all) 15 µm — Matches hyphal diameter

Chip Size

10×10 mm

Ports

16 reagent + 1 chow

Levels

4 branching levels

Compatibility

SU-8 / PDMS / Print

IP Coverage

MED Fungal-specific channel geometry (13–20 µm)
HIGH Living microfluidic device (integrated system)
MED Co-culture integration methods

H-Tree scaffold — 10×10 mm, 4 levels, 16 ports · channel width 13–20 µm, set against measured N. crassa hyphal geometry rather than adapted from PDMS conventions

Main channel (20 µm)
Branch channels (13–17 µm)
Reagent port
Chow reservoir

Five patent families — US provisional 64/087,071, filed June 10, 2026

Michael McKenna is sole applicant. The application is owned outright, with no institutional encumbrance.

# Family Description Strength
01 Metabolically orthogonal chassis Specific genotype + six-gene KO screening method HIGH
02 Optogenetic septal pore valve LOV2-Jα-HEX1 + SPA1-ePDZ constructs & flow control method HIGH
03 Living microfluidic device Chip + organism + scaffold as integrated system HIGH
04 H-tree scaffold, fungal dimensions Fungal-specific channel geometry (13–20 µm) novel over PDMS art MED
05 Six-gene chassis (Δcre-1 + Δwc-1) Requirement for 450 nm optogenetic activation in orthogonal host HIGH